| Popis: |
The grand challenges of our age such as climate change and food security are creating a greater need for cross-disciplinary scientific insight to guide understanding and inform decisions. The evolution of technology, with the introduction of big data approaches, cloud computing and access to storage and processing, is also changing the way data is accessed and used. As a result, researchers are demanding more detailed earth observations for assimilation into increasingly sophisticated models and simulations, and there is a clear trend toward larger computational undertakings across a loosely coupled set of geoscience disciplines. This, in turn, introduces the demand for access to diverse data sets from a broad range of scientific domains. These data have many different formats coming from distinct community cultures. Bridging scientific disciplines through the introduction of a broker or mediation capability has been demonstrated for a number of cross-discipline applications [Nativi, et al 2013]. The development and use of brokering middleware in a research environment is a relatively new development and the sustainability of such a capability has not previously been addressed. Broker software now has a level of maturity that requires addressing sustainment of the middleware, including the question of which business models could be adopted to best serve this need. The Working Group, in addressing the questions of sustainability, does not presuppose which business/revenue or hybrid model might be embraced to sustain this middleware, but rather examines five classes of business models for consideration. For each example, the strength and weaknesses in the context of long-term sustainment of broker middleware are discussed. Examination of alternatives leads to a prioritization among models with the recognition that no single model by itself is likely to provide the desired sustainment. We find a hybridized model incorporating aspects of three different business models over the lifespan of the brokering middleware, i.e. federally funded data facility guardianship in the establishment stage replaced or supported by a Consortium model and/or Software-as-a-Service as the broker matures, will likely provide the strongest model for sustainment. |
